Generally, concepts are forms of mental representation of a cognitive set of objects, whether natural, artificial or hypothetical (Eysenck 2012). We can say that these are sets of features related in meaning or function. There are two basic groups of concepts derived from the types of categorization: classical concepts and natural concepts (Eysenck 2012). The former are related to the theory of classical concepts and are based on an empiricist approach or a standard set theory assuming that it is possible to unambiguously determine the belonging of a given concept to a category where there are strict limits and binary attributes (Fehr and Russell 1991). Classical concepts are thought to be clearly defined, with clear definitional boundaries, and definitional structure. In turn, natural concepts are less precise, they have a more or less specific character. This type of concepts is derived from a trend called realism and is closely related to the theory of the prototype (e.g. Medin 1989). Natural concepts are dependent on changes brought about by experience and an impact of the situation. The construction of natural concepts explains the theory of fuzzy sets, and these concepts are elements of lay-representations of the world (Fehr 1988, 2005). This group includes emotional concepts. They will be understood as concepts that, to varying degrees, denote emotions such as love, liking; these are not words with emotional connotations such as dog, tree. The studies were focused on the concepts within the basic emotion spectrum, i.e. love is a basic emotion while liking, infatuation, passion are terms within love spectrum. These terms denote an emotion of love, they are emotional words but they do not carry an emotional connotation. Emotional concepts according to Niedenthal (2008, p. 587) are mental representations of categories, objects, situations, and activities. Emotional concepts contain information about the causes of emotions, i.e. situational factors potentially causing emotions, typical behaviors or activities corresponding to a given emotion, and internal subjective states typical for a given emotion, and its meaning (Niedenthal 2008, p. 588). In general, conceptualization of emotions is characterized by several properties: unfocused boundaries, existence of bipolar dimensions in some cases, category relativity, typology, existence of intercategorical relationships, linking with scripts and scenarios, hierarchical organization and rooting in more general theory of mind (Russell and Bullock 1986). A description of natural concepts, such as emotional concepts, requires a reference to natural/lay language (Wierzbicka 1992a, b). Language analysis enables description of mental representations in a form of concepts, because language is a part of the mind and simultaneously a phenomenon that constructs human mental and social capabilities (Wierzbicka 1992a). The best known models of emotional concepts formulated by scientists include dimensional, semantic units, prototype, and semantic network models (Niedenthal 2008; Niedenthal et al. 2006).
Another conception was formulated by Russell (1991, 2005) who proposed a concept of emotion as a script. Emotional concepts are events that have a causal and temporal structure. The concept of the script describes the structure of emotion concept, also taking into account prototype elements. Its structure includes key elements related to emotional situations more specifically such factors as: causes, beliefs, sensations, physiological changes, desires, activity, as well as vocal, and mimic expressions. The assumption about the prototype of affective concepts has been confirmed in many studies using material with different modalities, requiring recognition and naming of emotions or narratives. However, it has been shown that some emotion concepts are characterized by a distinct structural identity, e.g. they contain a list of features or unique specimens (according to the exemplar theory: Kroska and Goldstone 1996).
The aim of the current study is to present a new way of assessing the structure of emotion concept. And the example of this is a love concept.
Love: A Differentiated Emotion and a Differentiated Concept
Love is defined in the literature in a variety of ways: as a relationship, as an attitude, as an experience, and as an emotion (Fehr and Russell 1991). This complex emotion consists of affective experiences with positive valence, for instance joy, elation, sexual arousal, contentment, delight, a sense of happiness (Fehr 2006). Although positive valence dominates in this group of feelings, negative elements which may appear include anxiety, fear, dissatisfaction, jealousy, hatred, shame or guilt (Sternberg 1986). In the literature it has been emphasized that this emotion is not uniform in terms of the structure of affective feelings, the type of object to which the feeling is addressed, and the phase of relationship (Sternberg and Weiss 2006; Hendrick and Hendrick 2006). Furthermore, there are many other determinants influencing this complex emotion (Dion and Dion 1996, 2006; Hatfield et al. 2007; Schmitt et al. 2009). Researchers exploring romantic love have established that the differentiation of understanding and experiencing of love is determined by a huge number of factors (Shiota et al. 2010; Smith and Klases 2016; Sternberg and Weis 2006). A crucial role in this respect is attributed to the factors of gender, age, personality traits, individual experiences, and emotional maturity (Schmitt 2006). There are several theories of love proposed in psychological literature that have been comprehensively described, e.g. in the book entitled New psychology of love. One of the best known is the three-factor conception of love, formulated by R. Sternberg, assuming there are three components of love: intimacy, passion, commitment. These components evolve over time and are constantly changing (Sternberg 1986). Another conception by Sternberg defines love as an individual story that a person continuously builds in the course of development and experiences. It is a narrative concept containing basic generalized information about love. The nature of this narrative has a great impact on the quality and type of a specific person’s relationships. People construct different stories of love, and their abstractive versions are stored in memory forming a representation of affective experiences. The structure of this representation is hierarchical (Sternberg 1998, 2006).
Fromm’s (1956) conception of love, for instance, assumes that love is a character property, an individual’s attitude towards the world, and not to one person only. The theory by Hatfield (Lieberman and Hatfield 2006) highlights the existence of two basic types of love, passionate and companionate. Passionate love can be defined as a state of intense desire to unite with another person (Landis and O’Shea 2000). Companionate love is defined as sympathy and tenderness towards people with whom our lives are related. This kind of love is characterized by a very close friendship, sympathy, respect, mutual concern, and attraction (Fehr 2006; Hatfield and Rapson 1987, 1996; Kim and Hatfield 2004). Another conception of love formulated by Lee (Hendrick and Hendrick 1986, 2006) distinguishes six types of love, by reference to the Hellenic ideals of Eros, Storge, Ludus, Pragma, Mania, and Agape. The lay-conception of love described by Fehr points out that most people associate love with honesty, trust and caring, and rarely with addiction, sexual passion, or physical attractiveness. Fehr (2006) argues that lay-people have quite extensive knowledge about the concept of love. The lay-concept of love includes both the properties of companionate and passionate love. The most prototype features of love are trust, caring, and intimacy, whereas secondary features of love include staring at the other person, sexual passion or fast heartbeat. This means that the features of companionate love are considered to be more prototypical features of lay-concept love. Personal lay-conceptualization of love is associated with personality traits. For instance, people displaying high measure of protectiveness have a propensity for companionate love, while more dominant people for passionate love (Fehr 2005, 2006; Fehr and Broughton 2001; Fehr and Russell 1991).
As regards the concept of love, researchers have established that this concept has a prototype construction (Aron and Westbay 1996; Fehr 1988; Fehr and Russell 1991). Shaver and associates described the prototype structure of love taking into account words with emotional connotations. They have shown that love is basic/fundamental among emotional concepts. The structure of love concept includes factors preceding the occurrence of love, such as an atmosphere of openness, a sense of security, then an emotional mood: the relationship between people based on a profound feeling that they are liked, loved, need each other; it also incorporates activities: there is a very good compatibility between the people. As regards the manifestations of love, cognitive aspects are discussed as well, for instance difficulties with concentration, and attention focusing on a beloved person. Then, emotional aspects have been highlighted, e.g. a strong feeling of happiness and joy, and a sequence of behaviors expressing a need to be together. The components of the concept of love suggest that the concept of companionate love prevails in the whole structure of love (Shaver et al. 1987; Sternberg and Weis 2006). Elements of passionate love are definitely less intense in this structure, although they are also marked in the search for physical closeness, desire, and so on (Hatfield and Rapson 1987, 1996; Shaver et al. 1987).
The present research program was designed to explore the structure of the concepts representing the spectrum of love, taking into account linguistic material. This approach was based on the assumptions that language analysis can be of value in describing the semantic structure of emotional concepts (drawing on the thesis that “language is a tool used to express meanings”, as proposed by Wierzbicka 1996, p. 19; when we explore a language we refer to meanings which are a way of understanding something by a person).
Emotional Verbal Fluency
This paper addresses the problem of cognitive structure/internal organization of the emotional concepts from the viewpoint of semantic relationships expressed through emotional verbal fluency. Verbal fluency techniques belong to the classical psychological tools used in the diagnosis of cognitive functions, mainly in neuropsychology (Ardila and Ostrosky-Solí 2006). Verbal fluency techniques generally rely on spontaneous listing of words according to a criterion. Therefore, the subject is asked to name words representing a given category/criterion within a certain amount of time (usually 1 min) (Gawda and Szepietowska 2013, 2016). Apart from the standard applications of verbal fluency techniques, it was assumed that by using qualitative-quantitative analysis of words generated during emotional verbal fluency tasks, it would be possible to describe the organization and structure of emotional concepts. This approach is justified by the evidence showing that this technique enables effective exploration of verbal, episodic, and semantic memory (Goni et al. 2010). To our knowledge, few results based on the use of such a methodology have been published, and traditional indicators, such as the number of correct responses, errors, semantic switches, semantic clusters, phonemic switches, and phonemic clusters have been taken into consideration in the analyses; typical analysis usually consists in quantifying the total production of words (e.g. Abeare et al. 2017; Gawda and Szepietowska 2013, 2016; Rossell 2006; Sass et al. 2013; Wauters and Marquardt 2017).
Emotional/affective verbal fluency (EVF) is a novel and rarely used semantic fluency type which can take a variety of forms: positive versus negative, pleasant versus non-pleasant, joy versus fear, etc. (Gawda and Szepietowska 2013, 2016; Gawda et al. 2017; GSass et al. 2013). Furthermore, a new approach to examining the words generated by subjects has been proposed here. This analysis is focused on reconstructing a semantic network of words reflecting the structure of emotion concepts. Words generated in affective verbal fluency tasks reflect the knowledge available to the subject relating to particular emotions. Generated words are not uttered accidentally; on the contrary they are related semantically to each other, words form semantic clusters, and the content of these clusters can be described (Berto and Galaverna 2016). The contents of semantic clusters are interesting if we want to understand the structure of emotional knowledge. A semantic network composed of semantic clusters reflects the structure of the concept. This approach and the way of assessing emotional concepts has been inspired by research on the organization of semantic knowledge based on the previous studies, such as a network approach based on verbal fluency data in assessment: the organization of the animal category (Goni et al. 2010), behavioral differences between healthy subjects, patients with cognitive impairment (Lerner et al. 2009), multidimensional approach to verbal fluency in patients with frontal and right lobe damage (Schwartz and Baldo 2001), organization of concepts based on material derived from the letter fluency (Schwartz et al. 2003), semantic memory in patients with Alzheimer’s disease (Chan et al. 1993), semantic memory organization in verbal fluency test “Human Body Parts” in patients with schizophrenia (Berto and Galaverna 2016). Although some studies carried out so far have used the qualitative analyses of Animals and letter categories of verbal fluency tasks, none of them was focused on emotional categories of verbal fluency.
The aim of the present studies is not to search for the prototype properties of the concept of love but to describe the structure of the concepts within love spectrum, represented by the categories of liking, infatuation, love, and its differentiation on the basis of linguistic material. To our knowledge, this type of analysis using language has not been presented in the literature. In order to describe the structure of the concepts from love spectrum, EVFT tasks were used as we were inspired by the previous findings (Gawda and Szepietowska 2013).
Here is an overview of the two studies. The analysis of the structure of emotional concepts can provide insight to emotional functioning of a human being. The knowledge about the structure and functioning of emotion concepts is of great importance for understanding the human psyche. That is why the project aiming to describe the structure of emotion concepts has been undertaken. The whole project focuses on five groups of emotional concepts relating to basic emotions such as joy, sadness, fear, anger, and love (Russell 2005). These are concepts located in the taxonomic hierarchy between the superior (e.g. emotion) and the subordinate level (e.g. panic, fear). The part of the project reported in this paper refers to concepts within the love spectrum. Two studies have been conducted. The purpose of the first one was to analyze love spectrum concepts with use of the new technique i.e. emotional verbal fluency test (EVFT). The second was intended to demonstrate and confirm the psychometric properties of the new method for describing love concepts. The commonly applied verbal fluency techniques are based on a semantic or phonemic principle. In contrast, far less attention has been paid to emotional verbal fluency tasks. Psychology is still facing a need to develop appropriate tools to assess emotional aspects. There is a scarcity of methods allowing us to study the structure of concepts. The first study (Study I) reported here was designed to identify semantic clusters from words generated during love spectrum tasks, and to examine their relationship with the triangular theory of love by Sternberg; it was also intended to test reliability and validity of EVFT. The reliability of EVFT was assessed with test–retest (after 2 months) correlations and with inter-judge correlations. Construct validity is demonstrated by correlations of EVFT with other verbal fluency tasks and WAIS-R Vocabulary subtest, while discriminant validity is shown by correlations with Sternberg Triangular Love Scale (1997) and The State Trait Anxiety Inventory. Study II was performed to verify the semantic clusters identified in Study I, and examine the relationship between the semantic clusters using another conception of romantic love i.e. love attitudes proposed by Lee (measured by the Love Attitudes Scale). In the two studies, verbal intelligence was controlled; Vocabulary subtest scores were used in the screening procedure to qualify participants without impairments in verbal comprehension, and then, in correlational analyses with regard to reliability testing. The main research questions focused on what kind of semantic clusters could be identified from words generated during EVF tasks related to liking, infatuation, and love. The next questions were: what are the differences in identified clusters between tasks related to liking, infatuation, and love, and what are the internal relationships (within the categories of liking, infatuation, love) between semantic clusters? Then, we wanted to find out whether these semantic clusters are related to love concept in the Triangular Theory of Love, and love attitudes by Lee. Finally, we wanted to see whether EVFT is reliable in terms of inter-rater agreement and test–retest correlations, whether EVFT is characterized by acceptable construct and discriminant validity, i.e. whether EVFT scores correlate with other fluency tasks, verbal intelligence, and STAI scores.
Words generated by participants are semantically related and form semantic clusters. It is assumed that clusters reflect personal concept (lay concept) of liking, infatuation, and love. Emotional verbal fluency technique reveals lay-conceptualization of concepts within the spectrum of love. The components of the lay concept of love are consistent with the triangular theory of love; they correspond to intimacy, passion, and commitment. All lay concepts in the love spectrum contain these elements, however, to the different degree. We expect that the lay concepts of love also contain other elements such as aesthetic, existential, and temporal, because love is related to personal values and one’s personality as well as motivations. The assumption of the hypothesis is based on the previous findings showing that lay-people listed many more features of love than scientists (Fehr 1988; Fehr and Russell 1991; Gawda 2008). Finally, we expect that lay-concept of liking, infatuation, and love contains an aspect identified in the theory of love proposed by Lee, i.e. Storge, Eros, Ludus, Mania, Agape, and Pragma.